My data.frame table consist of 3 variables (x,y and z) where each variable has 1000 units. I need to create 5 equal size strata according to one of the variable (let's say x) whereas units of x variable with a higher value have higher probability to be selected in a strata with a higher number (max strata number is 5). I've been trying different things so far and since I am fairly new to R i can't figure it out how to trick the R to produce me a table with those 5 strata containing values of my original data so that I could continue with a further analysis. Please help me! have tried Survival package (Stratified sampling) but it doesnt produce me results in the right way...i must have been doing sth wrong [[alternative HTML version deleted]]
On Wed, Jun 18, 2008 at 9:20 AM, Ana Kolar <annakolar28 at gmail.com> wrote:> My data.frame table consist of 3 variables (x,y and z) where each variable > has 1000 units. I need to create 5 equal size strata according to one of the > variable (let's say x) whereas units of x variable with a higher value have > higher probability to be selected in a strata with a higher number (max > strata number is 5).Could you be a little more precise about the desired distribution within the slices? For example, one way to solve your problem as specified would be to sort the data frame by x, and chop the data frame into five equal pieces. Would that be a valid solution to your problem? Hadley -- http://had.co.nz/